Letter from America: Innovative worries in the Age of Big Tech

Daron Acemoglu on the challenges of modern innovation for society and economic theory.

If there is one thing on which economists and the American public agree on, it is that innovation is good. For economists, innovation is both the engine of growth and the lifeblood of a dynamic economy. For the public, it is the special sauce for US exceptionalism.

Innovation’s important role is also the reason why the American public and regulators have been tolerant of the overwhelming dominance of the tech giants, such as Meta (previously Facebook), Alphabet (previously Google), Apple, Amazon, and Microsoft. These companies have reached sizes unprecedented by any other corporate structure in history, and policymakers and public intellectuals should have been alarmed about their size, power, and effects in and beyond their sectors. But this hasn’t happened, at least until recently.

Arguably, the chief reason for this is their innovativeness. The tech giants invest a huge amount in research and development, and have been at the forefront of many of the defining technologies of the late 20th century and early 21st century, most recently those related to artificial intelligence.

But what if not all innovation is created equal? What if the real damage that tech giants, and for that matter other large firms, create is by distorting the type of innovation an economy is engaged in?

There are a number of reasons why these questions may be more pertinent today than before.

To start with, a lot of the technology-related spending from tech giants is targeted at purchasing products and production methods developed by other companies. This creates at least three major problems for innovative dynamism. First, these deals increase tech giants’ dominance, for example, as the acquisition of WhatsApp and Instagram by Facebook illustrates. Facebook’s internal documents show that its top executives were indeed motivated by a desire to nip a potential competitor in the bud.

Second, many of these purchases are “killer acquisitions” — the technologies that the tech giants purchase are left idle rather than being integrated into the core products of the acquiring companies. Presumably, the reason why they are being purchased in the first place is that they would have otherwise turned into rivals.

Third, and potentially most damaging, the prospect of being acquired, rather than competing against some of the most powerful organizations in the world, is attractive to many startups and entrants. It also shapes their direction of innovation. When we think, for example, of the difficulties that would be faced by an app threatening Apple’s app store or Google’s Play Store ecosystem, we come to recognize that today’s environment creates a symbiotic relationship between startups that strive to be acquired and tech giants that can maintain their monopoly position by such acquisitions.

This symbiotic, but pernicious, relationship is not the only reason we should worry about innovation in the tech sector, however.

Although each one of these major companies is investing in some areas that could generate significant benefits for humanity (fighting climate change, detecting cancers, etc.), many of the research dollars are targeted at better methods of data collection and processing. Are we truly going to benefit when Google and Facebook can monitor our movements and preferences more faithfully, so that they can target more manipulative ads?

In fact, it is not obvious that all innovation in a monopolistic market always benefits consumers. New algorithms and data collection methods may be targeted at increasing a monopolist’s grip on the market or boost its ability to convince users to spend more time on its platform or buy products that it markets. Should society embrace a new technology that enables a platform such as Facebook or Google to find out more about the preferences, views, and biases of individuals, even those that they would like to keep private?

Innovations that have complex societal effects go beyond those that directly empower platforms and disempower consumers. The business model of the tech giants is broadly centred on algorithmic automation — the use of algorithms and machines instead of humans to perform a range of tasks.

Automation is not a new phenomenon. It has been going on rapidly at least since the beginning of the Industrial Revolution in Britain, starting mid-18th century. It’s also undoubtedly true that economic growth depends on automation, so that labour can be freed from more routine, manual tasks in agriculture and low-skill manufacturing to be allocated to new tasks, industries, and services (and also towards greater leisure). But automation also creates major distributional effects. My own work with Pascual Restrepo, for example, finds that between 50% and 70% of changes in the US wage structure between detailed demographic groups is accounted for by the rapid spread of automation technologies in factories and offices. Automation has also been a powerful force holding back average wage growth in the US.

Should we celebrate new automation technologies being rolled out by the US tech sector? The answer is not obvious. On the one hand, anything that increases the productive capacity of an economy creates benefits. On the other, fewer jobs, less wage growth, and greater inequality are socially costly. My work with Pascual Restrepo also argues that a focus on automation often comes at the expense of ignoring other productivity-enhancing investments in worker-friendly technologies, for example those that create new tasks in which workers can be centrally reinstated into the production process.

If so, the focus on algorithmic automation spearheaded by tech giants may have considerable costs, as well as benefits.

All of this points in the same direction. Innovation, though it continues to be essential for a dynamic market economy, has more complex effects, and this is doubly so for innovation coming from the tech giants. We should watch out for what the new algorithms, data collection methods, and algorithmic production techniques are doing. We should wonder whether the massive research expenditures from these companies are mostly for the benefit of US and global consumers, or more often targeted at solidifying these companies’ market position, to disadvantage rivals, and create benefits for them at the expense of the poorer segments of society.

And if innovation has more variegated effects on society, this is not bad news for economic theory. It means that we need to invest more in understanding the consequences of innovation and develop more of a focus on the type of innovation, not just its quantity. It also means that economic analysis may have a greater role to play in the regulation of what will become the biggest and most consequential sector of the economy.

Daron Acemoglu, 19 August 2022

References and further reading

Acemoglu, D. (2021). Harms of AI. MIT, August. https://economics.mit.edu/files/23149

Acemoglu, D. and Restrepo, P. (2018). The race between man and machine: Implications of technology for growth, factor shares, and employment. American Economic Review, 108 (6), 1488-1542.

Acemoglu, D. and Restrepo, P. (2020). The wrong kind of AI? Artificial intelligence and the future of labour demand. Cambridge Journal of Regions, Economy and Society, 13 (1), 25-35.

Acemoglu, D. and Restrepo, P. (2022). Tasks, automation, and the rise in US wage inequality. MIT, February. https://economics.mit.edu/files/22821

Cunningham, C., Ederer, F. and Ma, S. (2021). Killer acquisitions. Journal of Political Economy, 129 (3), 649-702.

Glick, M. and Ruetschlin, C. (2019). Big Tech acquisitions and the potential competition doctrine: The case of Facebook. Institute for New Economic Thinking Working Paper Series no. 104, October, https://doi.org/10.36687/inetwp104


Keynes, J. M. (1930). Economic possibilities for our grandchildren. In: Essays in Persuasion. Palgrave Macmillan, London, 2010. https://doi.org/10.1007/978-1-349-59072-8_25 or RES members can download the CUP edition from the RES website, journals

Zeira, J. (1998). Workers, machines, and economic growth. Quarterly Journal of Economics, 113 (4), 1091-1117.

Related resources